Re: linint2_Wrap Regriding

Hi Ping,
I would think that you want the high resolution T and PSL fields to look
like the low-resolution T and PS fields, right? linint2_Wrap is
basically a simple linear interpolation routine, fine for many
applications. The T and PSL T62 fields are likely relatively smooth,
while I am guessing the T62 solar radiation field is not. The resulting
interpolated high resolution output will thus likely have smoother T and
PSL fields, and noisier solar radiation fields.

Speaking to a regridding expert here, there are other interpolation
methods that you can use, such as bi-cubic sphere interpolation.
However, if you start out with a relatively smoothed field as your T62
PSL/T fields are, it's not going to matter whether you use a linear
interpolation or a bi-cubic sphere interpolation routine, the results
will look quite similar.

Hope that helps.. Adam

On 05/18/2012 10:02 AM, Ping Yang wrote:
> Dear Adam,
>
> Thank you for the advice. Previously I was thinking to subset before
> the interpolation which can save some computation for area that not
> interested. It is working now.
>
> One more question, the interpolation function linint2_Wrap seems work
> better with some variables like solar radiation but not good for air
> temperature and pressure, because I found that when I interpolated the
> air temperature and surface pressure, the interpolated result is the
> same as the original file just the region has been extracted, I think
> it is supposed to be more intervals(which is higher resolution). but
> for solar radiation it worked good.
>
> Do you (or some user here) have the same problem?
>
> Looking forward to hearing from you.
>
> Regards,
>
> Ping
>
> On Thu, May 17, 2012 at 12:45 PM, Adam Phillips <asphilli@ucar.edu
> <mailto:asphilli@ucar.edu>> wrote:
>
> Hi Ping,
> I think you are on the right track, but you do not need to subset the
> input array. (I am guessing that you are getting null values
> around the
> perimeter of your new grid.) When it comes to regridding one needs
> to make
> sure that the input latitudes and longitudes and the output
> latitudes and
> longitudes are in the same range, and flow in the same direction. You
> state you want your grid on negative longitudes, which is fine,
> but that
> means your input longitudes need to span the same range as well. So, I
> would recommend reading in the input data in its' entirety,
> flipping the
> latitudes, then use lonFlip to convert the longitudes from 0:360 to
> -180:180:
>
> pres=in->pres(:,::-1,:) ; ::-1 = flip latitudes
> pres = lonFlip(pres)
> printVarSummary(pres) ; lats should run from S->N, and lons from
> -180:180
>
> newlon = fspan(-84.025,-64.875,383)
> newlat = fspan(34.875,50.125,305)
> newlat@units = "degrees_north"
> newlon@units = "degrees_east"
> newsst=linint2_Wrap(pres&lon,pres&lat,pres,True,newlon,newlat,0)
>
> That should be it. If that doesn't work or if you have further
> questions
> please respond to ncl-talk.
> Best regards,
> Adam
>
>
> > Dear NCL user,
> >
> > I am doing interpolation from the PSD data(daily T62 Gaussian grid
> > (192x94)
> > 88.542N - 88.542S, 0E - 358.125E) to a 3 minute*3minute lon/lat
> regular
> > grid, I have encountered an issue with the bilinear
> interpolation with the
> > linint2_Wrap function (Interpolates from a rectilinear grid to
> another
> > rectilinear grid using bilinear interpolation, and retains
> metadata)with
> > NCL, because I've got invalid values(some of of the value are
> null). I
> > used
> > this code (I used a NCEP/NCAR reanalysis data, which is
> available by this
> > link
> >
> ftp://ftp.cdc.noaa.gov/Datasets/ncep.reanalysis.dailyavgs/surface_gauss/pres.sfc.gauss.2006.nc)
> > :
> >
> > ;*************************************************
> > ; interpolation.ncl
> > ;
> > ; Concepts illustrated:
> > ; - Interpolating from one grid to another using bilinear
> interpolation
> > ; use linint2_Wrap function for the retrieving all the metadata
> > ;create a new file to include the interpolated variable
> > ;
> > ;************************************************
> > load "/usr/local/lib/ncarg/nclscripts/csm/contributed.ncl"
> > ;load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/contributed.ncl"
> > ;change $NCARG_ROOT to /usr/local directly.
> > ;************************************************
> > begin
> > ;open file to write variable to(a new netCDF file)
> > out=addfile("pres.sfc.gauss.2006.interpolated.nc
> <http://pres.sfc.gauss.2006.interpolated.nc>","c")
> > ;************************************************
> > ; read in netCDF file
> > ;************************************************
> > in = addfile("pres.sfc.gauss.2006.nc
> <http://pres.sfc.gauss.2006.nc>","r")
> > ;************************************************
> > ; read in pres
> > ;***********************************************
> > ;Use a subscript for change the dimension of pressure and also
> subset the
> > variable
> > pres=in->pres(time|:,{lat|34.875:50.125},{lon|95.975:115.125})
> > ;so the newPres is with the latitude increasing
> > ;***********************************************************
> > ;newlon = fspan(-84.025,-64.875,383)
> > newlon = fspan(95.975,115.125,383);convert to a (0,360) instead of
> > (-180,+180)
> > newlat = fspan(34.875,50.125,305)
> > newlat@units = "degrees_north"
> > newlon@units = "degrees_east"
> > ;************************************************
> > ; interpolate to new grid use linint2_Wrap function
> > ;***********************************************
> > newsst=linint2_Wrap(pres&lon,pres&lat,pres,True,newlon,newlat,0)
> > ;*************output variable to new file****************
> > out->pres=newsst
> > end
> >
> > Actually I need the domain range is from lon(-84.025,-64.875,383),
> > lat(34.875,50.125,305) in the lon/lat regular grid, Can I ask
> you what
> > will
> > be the correct range in the Gaussian grid lon 95.975:115.125? I
> tried
> > several times, it seems that I am confused by the translation,
> >
> > Am I use the right function for this interpolation? I will be
> appreciated
> > if you can point out where is the problem ?
> >
> > Best Regards,
> >
> > Ping
> >
> > --
> > Ping Yang, Ph.D.
> > CUNY Environmental Crossroads Initiative
> > Marshak Science Building - Suite 925
> > The City College of New York - CUNY
> > 160 Convent Avenue, New York NY 10031
> > Phone: 212-650-5769 <tel:212-650-5769>
> > Fax: 212-650-7064 <tel:212-650-7064>
> > _______________________________________________
> > ncl-talk mailing list
> > List instructions, subscriber options, unsubscribe:
> > http://mailman.ucar.edu/mailman/listinfo/ncl-talk
> >
>
>
> __________________________________________________
> Adam Phillips asphilli@ucar.edu <mailto:asphilli@ucar.edu>
> NCAR/Climate and Global Dynamics Division
> P.O. Box 3000 (303) 497-1726 <tel:%28303%29%20497-1726>
> Boulder, CO 80307-3000
> http://www.cgd.ucar.edu/cas/asphilli
>
>
>
>
> --
> Ping Yang, Ph.D.
> CUNY Environmental Crossroads Initiative
> Marshak Science Building - Suite 925
> The City College of New York - CUNY
> 160 Convent Avenue, New York NY 10031
> Phone: 212-650-5769
> Fax: 212-650-7064
>
>
>
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk

-- 
______________________________________________________________
Adam Phillips                                asphilli@ucar.edu
NCAR/Climate and Global Dynamics Division       (303) 497-1726
P.O. Box 3000				
Boulder, CO 80307-3000    http://www.cgd.ucar.edu/cas/asphilli

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Fri May 18 11:06:12 2012